使用hechman MLE方法回归为何不出结果？而是下面这种情况：但换twostep就可以了，是不是数据问题？多谢指教

heckman cd lnVA lndistcap colony comlang lngdppc lngdp lnw i.year , select( lnVA colony comlang lndistcap  lngdppc lngdp )

Iteration 0:   log likelihood =  167.04965  

Iteration 1:   log likelihood =  264.75093  (not concave)

Iteration 2:   log likelihood =  277.73698  (not concave)

Iteration 3:   log likelihood =  278.61427  (not concave)

Iteration 4:   log likelihood =  278.83732  (not concave)

Iteration 5:   log likelihood =  278.92508  (not concave)

Iteration 6:   log likelihood =  278.95561  (not concave)

Iteration 7:   log likelihood =   278.9776  (not concave)

Iteration 8:   log likelihood =   278.9977  (not concave)

Iteration 9:   log likelihood =  279.02301  (not concave)

Iteration 10:  log likelihood =  279.05254  (not concave)

Iteration 11:  log likelihood =  279.06646  (not concave)

Iteration 12:  log likelihood =  279.08815  (not concave)

Iteration 13:  log likelihood =  279.10855  (not concave)

Iteration 14:  log likelihood =  279.52055  (not concave)

Iteration 15:  log likelihood =  279.69876  

Iteration 16:  log likelihood =  286.45945  (backed up)

Iteration 17:  log likelihood =  291.71001  

Iteration 18:  log likelihood =  298.65751  

Iteration 19:  log likelihood =  301.06298  (not concave)

Iteration 20:  log likelihood =   304.1646  (not concave)